Tom Mueller wrote:
Thank you for the reference. In looking at Shawn's final reply on this thread listing all of the advantages of having separate source packages, I'm wondering about how this would read if "locale facet" or "development facet" where substituted for "source" in them?
...
I'm still curious as to what determines whether something should be a facet or a separate package.

Clearly, we are heading along the path that locales should be facets. And the conclusion of that thread was that source should be separate packages. So is there a qualitative difference between these? Or is it just quantitative? Source is too much data? Not enough users are interested? etc. And if quantitative, when do the numbers justify using separate packages?

The key differences are that:
* source will be used by a small minority of users
* is not integral to the operation of the software
* has different administrative, provisioning, and process needs
* operates on a different publishing cycle for our current processes

The packages are primarily for delivering software to the user, and delivering the source is an ancillary need.

Also, I don't believe that the same group responsible for delivering a binary release to a repository (OpenSolaris RE) will be aware of or be involved in the source delivery process. However, I could be wrong.

Cheers,
--
Shawn Walker
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to